Characterizing Electrochemical Gas Sensors with Impedance Measurements

Gas sensing is a topic of broad research interest. Our page on tunable diode laser absorption spectroscopy (TDLAS) shows a good example of application. That example considers an optical investigation to quantitatively assess gas content, achieving extremely low detection limits. In this short blog post, we focus instead on the electrochemical (EC) technique, which features a fast, simple, and low-cost sample preparation, and look at the role played by impedance measurement.

We choose an EC sensor (MQ-3 alcohol gas sensor, Digikey part number SEN-08880-ND) such as those found in common breathalyzers. The sensor measures the exhaled alcohol gas content based on the impedance change of the SnO₂ sensing material. We connect the MFIA Impedance Analyzer to a 6-pin MQ-3 as illustrated in Figure 1. The left and right pins are connected to the Aux output providing the DC voltage to the 30.4 Ohm resistive heating coil (also measured with the MFIA in a separate experiment). The remaining 4 pins are connected to the SnO₂-coated gas sensing layer. This arrangement creates a standard 4-terminal impedance measurement configuration on the MFIA.

The working principle of SnO₂ is similar to the polymer 'electrolyte' used in the EC humidity sensing experiment discussed in a previous blog post. SnO₂ is a wide bandgap semiconducting material, however, so its impedance (and the effective carrier concentration) is more prone to temperature changes. As it also has almost no diffusion (Warburg) hindrance, the impedance measured at low frequencies or even at DC might be enough to tell a meaningful story (by contrast, we clearly observed unwanted parasitic inductance above 2 MHz). After 12 hours of heating, the equivalent series resistance (ESR) falls from ~500 kOhm to below 50 kOhm; the parallel capacitance remains less affected at ~1 pF level, as shown in Figure 2. This trend is likely to continue as heat dissipation has not yet reached a steady state. If more rigorous testing is required, the MF-PID option can be used with another temperature sensor, to make sure that the DC voltage from the Aux output can be adjusted accordingly to prevent overheating.

For the purpose of this blog post we decided to perform the measurements without heating, so that the impedance baseline drifts very slightly from 141.9 kOhm to 142.6 kOhm (0.5%) due to the sensor's temperature change. Figure 3a shows the data taken immediately after drinking a 330 ml bottle of beer. By breathing on the sensor, the impedance increases by 26.1 kOhm (18%) with a recovery time constant of 9 s. A second measurement carried out 10 minutes later shows that an impedance change is still clearly visible, although it is weaker at 4.5 kOhm (3%) and is associated with a similar recovery time. If the observed impedance change is linearly proportional to the alcohol gas content, we can conclude that the latter also decreased by a factor of 6. Given that we found the sensor to have a resistance of  ~140 kOhm and a capacitance of 1 pF, the RC time constant is below 1 us; the recorded recovery time can thus be mostly attributed to the alcohol gas desorption process.

Impedance measurements can play an important role in daily life, an example being that of breathalyzers based on EC alcohol gas sensors. In this blog post we found that the exhaled alcohol gas level can drop fast, namely by a factor of 6 after a 10-minute interval between measurements. Nevertheless, to determine the absolute gas content accurately it will be necessary to refer to the calibration curve in the sensor's datasheet for a specific temperature and humidity level.

If you're curious about the capabilities of the MFIA Impedance Analyzer, get in touch to set up an instrument demo.